09-23-2008 01:35 AM - edited 03-03-2019 11:39 PM
DOH-GSR-PE2#sh int g2/0/1 controller
GigabitEthernet2/0/1 is up, line protocol is up
Hardware is GigabitEthernet, address is 0019.55e9.f4cd (bia 0019.55e9.f4cd)
Description: ***Connected to OLD_4500-Gig4/47***
MTU 1500 bytes, BW 1000000 Kbit, DLY 10 usec, rely 255/255, load 34/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full Duplex, 1000Mbps, link type is force-up, media type is LX
output flow-control is on, input flow-control is on
ARP type: ARPA, ARP Timeout 04:00:00
Last input 00:00:00, output 00:00:00, output hang never
Last clearing of "show interface" counters 00:22:24
Queueing strategy: fifo
Output queue 0/40, 58350 drops; input queue 1/75, 33 drops
Available Bandwidth 1000000 kilobits/sec
5 minute input rate 115415000 bits/sec, 28397 packets/sec
5 minute output rate 134713000 bits/sec, 27906 packets/sec
35908851 packets input, 17612788274 bytes, 0 no buffer
Received 28758 broadcasts, 0 runts, 4518126 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 watchdog, 67269 multicast, 0 pause input
35645315 packets output, 21985400017 bytes, 0 underruns
Transmitted 2596 broadcasts
0 output errors, 0 collisions, 0 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier, 0 pause output
0 output buffer failures, 0 output buffers swapped out
Above is the output of show interface command. In the output we see some output queue drops with very less queue.
Output queue 0/40, 58350 drops; input queue 1/75, 33 drops
What could be the reason for this act?
09-23-2008 01:53 AM
Hello!
One card queue is smaller than the other one. As I see there is a lot of traffic on the interface so, the lower queue size is dropping packets as he has not room for them. This is a usual problem that I saw mostly when connection an old card to a new server which support a high throughput and is having a lot of traffic. In my case the solution was to replace the hardware.
Maybe somebody has a different approach for this issue?
Cheers,
Calin
09-23-2008 03:59 AM
Input queue drops usually indicate the device couldn't process the input stream fast enough, and sometimes requires analysis and effort to elminate.
Output queue drops usually indicate (transient) congestion, and for such, is quite normal.
Since you only had 33 input drops for 35,908,851 packets input, it shouldn't be of concern.
You had 58,350 output drops for 35,645,315 packets output, which isn't real bad at about .16%. For gig speed on a LAN, and allowing for a BDP based on about 1 ms latency, you could pretty safely double your output queue size to 80 and see if that decreases your drop percentage.
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: